The allure of hedge funds is that they promise to earn good returns, whether the market is rising or falling. Reserved for the wealthiest investors, these exotic funds vary in their investment styles and returns, but they all share a tendency toward illiquidity, opacity, hefty fees and high investment minimums.
Those barriers keep ordinary investors from investing in hedge funds. But the introduction of low-cost clone hedge funds, constructed from exchange-traded funds (ETFs) by using quantitative investing techniques, may let Main Street enjoy hedge fund-like results.
It is too early to tell if hedge fund replication will democratize the benefits of hedge funds. But it is clear that experiments are contributing to a wider effort to improve asset allocation and portfolio optimization. Merely using standard asset classes is no longer good enough.
WHY PAY TWO AND 20?
About 6,000 active hedge funds now manage an estimated $1.5 trillion with varying investment styles and returns. Hedge funds traditionally charge clients 2% of assets and 20% of profits.
Hedge fund clones are relatively new. Two professors at London's Cass Business School, Harry Kat and Helder Palaro, planted the seeds for democratizing hedge fund investing with their 2005 paper, Who Needs Hedge Funds? A Copula-Based Approach to Hedge Fund Return Replication.
"Hedge fund return replication is about generating hedge fund-like returns by mechanically trading traditional asset classes," they wrote. Computers can do the work, with the choice of asset classes and trading rules based on regression analysis using historical data on aggregate hedge fund returns. The technique draws on work done in the early 1990s by Nobel Prize laureate William Sharpe, who modeled the performance of various types of mutual funds.
It took a few years before anyone tried training computers to create mutual funds that imitated hedge funds. A few existing mutual funds used basic hedge fund techniques. Diamond Hill Long-Short-Fund employed short selling, for example, and the Calamos Market Neutral Income Fund used convertible arbitrage, but these were actively managed funds.
BlackRock now offers iShares Diversified Alternatives Trust (ALT), which is an ETF based on an active strategy of seeking out market mispricings. Standard & Poor's has an index that dynamically allocates between the S&P 500 and a volatility hedge.
The replication movement lost steam during the last downturn, as hedge funds themselves lost some of their aura of invincibility. Although the average fund lost only 19% in 2008, or about half that year's drop in the S&P 500, a number closed. Many investors were stunned to learn that their high-cost, sophisticated strategies had failed to hedge.
THREE CLONING TECHNIQUES
Still, a few investment firms pushed ahead and brought out the first true clones. The pioneer in these efforts was IndexIQ, a firm that now offers ETFs based on hedge fund strategies. This firm and others build clones by using three basic methods: factor replication, mechanical trading strategies and distribution replication.
• Factor replication. This technique uses regression analysis to identify the factors-stock returns, price/earnings ratios, credit spreads, momentum measures-of hedge fund performance. The next step is to create similarly performing portfolios of traditional asset classes. IndexIQ's multi-strategy and macro-tracker funds use ETFs.
• Mechanical trading. This method uses computer-based rules that mimic the trading patterns of actual hedge fund managers. IndexIQ's merger/arbitrage fund sells the acquirer's stock and buys the target's stock when a deal is announced. After studying thousands of past acquisitions, IndexIQ built rules to help identify deals where this strategy is most likely to pay off.
• Distribution replication. This strategy attempts to construct investments that meet general statistical criteria, without attempting to reproduce month-by-month returns. For example, Kat collaborated with Aquila Capital in Germany on the Statistical Value Market Neutral Fund, which aims for annual returns of 10% or more, 7% volatility and zero correlation with the S&P 500.
Factor replication is the oldest and still most popular method of hedge fund replication, but this is still a rich area for research, and techniques are continually evolving. Some types of hedge funds are easier to clone than others. Long/short equity funds, for example, are easier to replicate than niche strategies, which have less general market exposure.
The growth of ETFs has made it easier to replicate hedge funds, since clones can be built by combining investments in multiple ETFs. Specialized ETFs provide extra raw material to replicate hedge fund performance more closely.
IndexIQ, the leading marketer of ETFs that replicate hedge fund strategies, currently offers four funds. They are the broad based IQ Hedge Multi-Strategy Tracker (QAI), the more targeted IQ Hedge Macro Tracker (MCRO), the IQ Merger Arbitrage ETF (MNA) and the IQ Alpha Hedge Strategy mutual fund (IQHIX/IQHOX), which is similar to QAI.
Credit Suisse markets two exchange-traded notes (ETNs): Merger/Arbitrage Exchange-Traded Notes (CSMA), which uses mechanical trading, and the Long/Short ETN (CSLS), based on factor replication. ProShares recently announced plans for an ETF that will use factor replication to match the performance of the HFRI Composite Hedge Fund Index.
Despite impressive progress, questions remain about the hedge fund cloning process. Perhaps the biggest challenge is the limited usefulness of historical data to predict future results. Returns, volatilities, correlations and other investment relationships change over time, which can reduce the effectiveness of hedge fund cloning techniques.
Also, the deft (or lucky) timing moves of the most successful managers may defy mimicry. Much is still unknown simply because hedge fund replication has such a short track record. But some data is available.
Swiss researchers Wallenstein, Tushschmid and Zaker analyzed performance data for 19 replication funds and indexes (which typically precede fund introduction) from March 2008 to May 2009. Returns differed for the various replication approaches, ranging from slightly positive to almost as negative as the 30% loss for the S&P 500. On the positive side, the researchers found hedge fund replication products to be competitive with their actively managed counterparts.
We can perhaps learn more about what to expect from replication by looking at performance history for actively managed hedge funds. After all, this is the performance that replication is attempting to match. "A Smoothie," on page 99, shows that hedge funds helped smooth out performance during the worst of the market crash, besting the S&P 500, commodities and real estate in 2008. But they by no means insulated investors from stock market losses in 2008; hedge funds lost 19% that year.
Over the most recent five-year period from 2006-2010, hedge fund returns were double the commodity and real estate returns and triple the returns from stocks. However, caution may be in order when interpreting these hedge fund results because of reporting biases noted by Ibbotson and other researchers.
In a March 2010 study of hedge fund performance, Ibbotson researchers Roger Ibbotson, Peng Chen and Kevin Zhu estimated that for the 10-year period 1995-2009, hedge funds averaged returns of 7.63% a year, after expenses, compared with 8.04% for the S&P 500. The authors adjusted their results to correct for biases in hedge fund reporting.
Should we expect replication funds to match hedge fund returns? To answer that question, we need to break hedge fund returns into their components.
The Ibbotson study estimated that the 7.63% return could be split into alpha of 3.01% attributable to fund manager skill and beta of 4.62% linked to systematic factors like the equity premium and credit spreads. Active management added value in every year except 1998.
Clones have the advantage of much lower fees. Today's ETF-based replication funds charge only about 0.75% a year, compared with an effective total cost (expense charges plus performance fees) that Ibbotson estimated at 3.78% for active hedge funds.
Since the cost advantage for clones roughly equals the estimated alpha generated by active hedge fund management, we might expect the two approaches to produce comparable returns after fees. That's how it appears to play out in the real world.
IndexIQ's CEO Adam Patti says his firm's multi-strategy ETF (QAI), "has produced results in line with the HFRI Fund of Funds Index." He notes that many investors, including fund-of-funds managers, use QAI as a low-cost core fund alongside actively managed funds, an indication that replication can match active management averages.
So it appears that hedge fund clones, like hedge funds, can produce near-equity returns with bond-like volatility, depending on the particular strategy being cloned and assuming lower costs. That makes hedge fund replication a potentially strong tool for optimizing portfolios for financial advisors and their clients, adding value to an asset allocation strategy.
TOWARD A FINANCIAL GENETICS
A new product called factor-based indexes could help spur hedge fund replication. Such indexes are basically new asset classes that, along with traditional asset classes, can serve as the raw material for replication techniques. Russell Investments and Axioma, an innovator in risk analysis and portfolio optimization products, have launched a series of these indexes, which track performance-related factors like momentum, volatility, liquidity, leverage and beta .
Intended for both investing and hedging, the factor indices are designed as a highly tradable basket of stocks that closely tracks an element of risk and minimizes exposure to other risks. Turnover, transaction size and the number of companies will be strictly limited to keep cost down. "There is no substitute for a skilled fund manager, but factor-based strategies are a natural for those who want to include hedge fund attributes in their portfolios," says Sebastian Ceria, Axioma's CEO.
MSCI Barra also offers factor-based indexes, providing exposure to momentum, volatility, leverage, value and earnings yield. The company has licensed the indexes to BlackRock's iShares in the United States and to Deutsche Bank's db x-trackers in the United Kingdom.
Hedge fund replication, however, should be seen as a way to add value to overall asset allocation strategies. Instead of trying to clone a long/short strategy or a merger/acquisition fund, we should think in terms of building new investment options with specific performance characteristics. In fact, much of Kat's current work focuses on these broader asset allocation issues.
For example, money managers could use ETFs to combine factor analysis and trading rules, building investments with attractive returns, low volatility and low correlation with stocks-without necessarily replicating any particular type of hedge fund. Once they're out of the experimental stage, it might be feasible to build whole portfolios around such investments instead of confining them to the supporting role of alternatives.
Joseph A. Tomlinson, FSA, CFP, is an actuary and financial planner based in Greenville, Maine. He devotes a significant amount of his time to researching and writing on planning topics.